Journal of Clinical and Diagnostic Research (Feb 2021)

Appraisal of Inter-rater Agreement among Assessment Scales and Retrospective Analysis of Validated Reports of Cutaneous Adverse Drug Reactions at Tertiary Care Hospital in Eastern India

  • Mukesh Kumar,
  • Pramod Kumar Manjhi,
  • Shruti Singh,
  • Soni,
  • Dheeraj Kumar Singh,
  • Sanmita Deo

DOI
https://doi.org/10.7860/JCDR/2021/46700.14547
Journal volume & issue
Vol. 15, no. 2
pp. FC17 – FC17

Abstract

Read online

Introduction: Cutaneous Adverse Drug Reactions (CADRs) share significantly to Adverse Drug Reactions (ADRs) comprising 10%-30% of all ADR reporting in India. Multi Drug Therapy for Leprosy (MDT-L) and antimicrobials contribute remarkably to the overall CADRs burden. Aim: To show distinctive pictures of CADRs profile and to assess inter-rater agreement of assessment scales among study populations. Materials and Methods: A retrospective analysis was done for 245 CADRs reported from March 2018 to March 2020. Cohen kappa statistics was applied for inter-rater agreement study for causality (World Health Organisation-Uppsala Monitoring Cente {WHO-UMC} Scale and Naranjo’s Algorithm), severity (Hartwig and Siegel scale) and preventability assessment (Modified Schumock and Thornton scale). Results: CADRs contribute 45.54% of total ADRs reported during study period. Male (60.41%) and age group 21-40 years (22.45%) were predominant sex and age group, respectively. Multidrug therapy for leprosy (51.83%) was the most common offending agent and hyperpigmentation (20.82%), dryness (13.1%), and both (11%) were the most prevalent CADRs. Causality of WHO– UMC Scale was higher with ‘Possible’ than ‘Probable’. Whereas, ‘Probable’ was maximally found with Naranjo’s Algorithm. Severity assessment showed maximum ‘mild’ cases i.e., 66.53% (manual) and 69.8%% (app). Preventability assessment depicted mostly ‘Definite’, 66.53% (manual) and 85.71% (app). Inter-rater agreement study showed ‘Substantial agreement’ for WHO-UMC Scale (K=0.678) and Naranjo’s algorithm (K=0.820), when manual vs app ratings were compared. ‘Almost perfect’ for severity assessment (K=0.893) and ‘Moderate’ for preventability assessment (K=0.434) were noticed. ‘Fair’ agreement was observed when manual (WHO-UMC scale) vs manual (Naranjo’s algorithm) were compared with K=0.290 and also, in app (WHOUMC scale) vs app (Naranjo’s algorithm) with K=0.319. Conclusion: CADRs were most prevalent among ADRs which have a distinctive picture in eastern India. WHO- UMC scale and Naranjo’s algorithm concluded significant differences in causality with only ‘fair’ agreement between them. Severity and preventability assessment done by manually little varied in their results with pharmvigill app and is still more reliable and popular.

Keywords